claritystorm/fda-faers-drug-adverse-events
FDA FAERS Drug Adverse Events 2023 One-time dated snapshot. Reconciled 2023 FAERS extract with 1,541,888 retained cases and seven related tables; supplied 2023 deletion lists applied. Verified coverage: 2023 Q1–Q4. Records: 1,541,888 case rows; seven related tables. This repository contains a 1,000-row public sample, not the full package. A sample does not establish complete historical coverage. Limitations Historical 2023 Q1–Q4 quarterly-extract cohort, not a… See the full description on the dataset page: https://huggingface.co/datasets/claritystorm/fda-faers-drug-adverse-events.
FDA FAERS Drug Adverse Events 2023
One-time dated snapshot.
Reconciled 2023 FAERS extract with 1,541,888 retained cases and seven related tables; supplied 2023 deletion lists applied.
Verified coverage: 2023 Q1–Q4.
Records: 1,541,888 case rows; seven related tables.
This repository contains a 1,000-row public sample, not the full package. A sample does not establish complete historical coverage.
Limitations
- Historical 2023 Q1–Q4 quarterly-extract cohort, not a current database or a count of events first occurring in 2023. No later quarters or deletion lists are included.
- Keeps the maximum case version observed per case within these four extracts. The union of their supplied deletion lists excludes 3,103 previously retained cases and their related rows.
- All child records match the retained DEMO primaryid, caseid and extraction quarter. Reconciliation removed 2,819,928 child rows from superseded reports.
- DRUG primaryid plus drugseq identifies a drug group, not a unique row: FDA permits multiple dosing/route records. All 623 repeated groups (636 additional rows) are preserved. drugrecord_id is unique only within this release.
- INDI.indidrugseq and THER.dsgdrugseq link to DRUG.drug_seq within primaryid. Join through distinct drug groups or aggregate first to avoid multiplying indication and therapy rows. Joining multiple child tables can also multiply records.
- Reports do not establish causation, incidence rates or comparative drug safety. Reactions are report-level and cannot automatically be attributed to an individual drug.
- Existing cleaned values, partial dates and substantial missingness are preserved. eventdt and mfrdt remain numeric source-date fields; drug rechallenge is stored in rechal. Validation covers packaging and case/drug relationships, not clinical interpretation.
Package upload: 2026-09-19. Listing reviewed: 2026-09-19. These dates do not extend source coverage. This review did not refresh the data.
Existing package tables
Counts are per table, not independent events. Full field names and types are on the dataset page.
Inspect the public CSV
import pandas as pd
df = pd.read_csv("sample_1000.csv")
print(df.shape)
print(df.columns.tolist())
print(df.head())Availability
The dated snapshot is available for a one-time $149 purchase, subject to its documented limitations and license terms. Future updates are not included.
New annual update subscriptions and bundle purchases are paused. Existing customers retain their purchase download access and can contact support@claritystorm.com for assistance.